Prevent software from running on new computers or virtual machines.

too fast processors (Pentium II and later).

You mean when they fixed the 1=.99999999999999 error ? You can do better than that. Once I get my state of the art PC totally together I can send yo u a program that will not run on anything with a math coprocessor, that mea ns nothing better than a 486SX. Also didn't they make math coprocessors for the 386 ? They were separate and maybe could be turned off in BIOS. If not you cod probably just remove them because what I was was in a separate soc ket. Or were those to take a 486SX to a DX ?

It has been a while. I don't even remember a third of the cars I've had or places I lived anymore.

Well yes but not. I got one. It is not that bad old, takes plenty of RAM, was running Win 10 until I fixed that.

Bit still, it is not for sale.

n too fast processors (Pentium II and later).

r than that. Once I get my state of the art PC totally together I can send you a program that will not run on anything with a math coprocessor, that m eans nothing better than a 486SX. Also didn't they make math coprocessors f or the 386 ? They were separate and maybe could be turned off in BIOS. If n ot you cod probably just remove them because what I was was in a separate s ocket. Or were those to take a 486SX to a DX ?

r places I lived anymore.

IIRC I used a 286 XT with maths copocessor. It was in a separate box!

NT

on too fast processors (Pentium II and later).

ter than that. Once I get my state of the art PC totally together I can sen d you a program that will not run on anything with a math coprocessor, that means nothing better than a 486SX. Also didn't they make math coprocessors for the 386 ? They were separate and maybe could be turned off in BIOS. If not you cod probably just remove them because what I was was in a separate socket. Or were those to take a 486SX to a DX ?

or places I lived anymore.

Yeah, 286 and 386 had separate chips for floating point math which was much faster than software emulation. It didn't really speed up spread sheets t hough as they have a lot of other computations having to do with addressing and accessing the info. I guess a spread sheet is a bit like an interpret ed language.

The 486 didn't need a floating point chip as it was already in there, but o nly in the DX version. The SX version turned it off. They also offered a floating point coprocessor for the SX version, the 487SX which was actually a 486DX with an extra pin. lol

--

  Rick C. 

  --+ Get 1,000 miles of free Supercharging 
  --+ Tesla referral code - https://ts.la/richard11209

snipped-for-privacy@gmail.com wrote in news: snipped-for-privacy@googlegroups.com:

486 was the first with it integrated on the same die as the main cpu. Now ubiquitous and "CPU" means so much more that used to be on the mobo. The 486SX was full bore 486 dies that failed on the mathco in die test, and they turned it off and called it an SX and sold them instead of losing what was at that time almost 25% of their die burns. They shut down ALCOA, who was making them right here in the USA for a while. 1500 employees... ZAP!

It's a common myth that alternate testing on the same chip was actually fai led devices. Someone from Xilinx once explained to me that it was not econ omically practical to worry with doing it that way. Instead if they wanted the fast speed grade, they tested to the fast speed grade. If they wanted slow speed grade the tested to the slow speed grade. They never retested chips to see if they would work at another speed grade.

I don't know where you got your 25% figure, but if they were rejecting 25% of the die because of failed floating point units, they likely would have h ad a 75% failure rate from the chip as a whole. So I doubt that is reality .

I expect they were doing what Xilinx does, they test the functions on the c hip they are interested in for that run.

--

  Rick C. 

  -+- Get 1,000 miles of free Supercharging 
  -+- Tesla referral code - https://ts.la/richard11209

10) ON ERROR, GOTO END 20) END:

Great suggestion!

You could also check for a dead CMOS battery. Because by now, it's pretty much a given that the battery is well past its freshness date. :)

Another idea is to try and install a Kaypro printer driver on it. If it won't do that, (or even if it will), the computer is too new. :)

I suspect there are not many people here old enough to get that joke.

I don't know, but my NEW full-size dot-matrix printer has a USB port. Does that count?

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

It is a known, Intel acknowledged fact that a 486SX were 486 DX targeted dies that failed in test on the MCP.

So, your common myth must refer to something other than my actual fact.

mpm wrote in news: snipped-for-privacy@googlegroups.com:

Let me put my proprietary RAM into my Compaq overpriced POS PC and see if it will load 'Em up.

I would love to see that information. Where did you see that?

--

  Rick C. 

  -++ Get 1,000 miles of free Supercharging 
  -++ Tesla referral code - https://ts.la/richard11209

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

It was not as in the rumor article about filling prders.

The ENTIRE SX chip line was 486 exact same die with the MCP disabled.

Why anyone would have bought a CPU for a mobo that will never have a coprocessor I do not know.

So you don't actually have any real info about that? I figured as much.

The 487SX was the same chip with the floating point enabled.

To save money since there were few applications where the floating point would actually speed up the processing very much. Duh!

The 486 was a fast part for the time. The AMD part was even faster. I think they cranked the clock up faster than Intel did and made a lot of money on it.

I forget. Was it the 486 that got AMD sued and they had to clean room design the microcode? Or was that the 386?

--

  Rick C. 

  +-- Get 1,000 miles of free Supercharging 
  +-- Tesla referral code - https://ts.la/richard11209

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

Fuck you. I read it locally in San Diego way back when they closed the Alcoa fab plant.

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

Not true by the time the 486 was in play. Duh yourself, twerp.

Engineering groups and CAD. AutoCAD started on a 286, so you can bet they coded full floating point by the time the 486 hit. That was about the most stupid thing I have ever heard you spout yet (the day is young).

Back with the 286, one could save by NOT buying the FPU. But even Lotus123 benefited so it was commonly included from the get go by any company wanting things to go a bit quicker.

That tends to make it only work with certain archane video cards unless you stick to the most basic poxy low resolution IBM CGA modes.

NECs V20 and V30 could both emulate 8080 OK. Intel was not amused. My first home PC compatible was an Epson PC using that chip.

It also had innovative bit twiddling instructions - handy for CPU driven memory mapped graphics drivers.

--
Regards, 
Martin Brown

Makes sense. transistor that do not fail a nondestructive test at a certain current or whatever get a higher number. Hard drives out of the oven are f ormatted to higher and higher capacities until the error rate becomes unacc eptable so then they ship them at a capacity they can handle.

In a way they take what they can get. I say like if you want a reliable har d drive and they offer say a 2TB and a 1TB, take the 1TB.